Master Theses Proposal Applications of Graph Transformation Systems to the specification of web applications

The World-Wide Web (WWW, W3) is a well-known distributed information system, that has grown during the nineties, and it is presently the most common way to publish information on the internet (see [W3]). Despite its tremendous expansion and impact on the computer industry, the basic WWW data model is still very poor (namely a flat collection of documents called pages, with references between them), and does not offer explicit support for the development of large information systems. This leads to common problems, like the broken link problem (since it is difficult to track all the references to a certain page, changes in the page structure can result in dangling references inside the WWW), or the lost in hyperspace problem (since the WWW has a flat structure, people often get lost while browsing through the web). In such cases, it is useful for the authors and/or maintainers of web pages, to have some tools that support the development and maintenance of information in the web. For example, web site managers can be helped to move or delete pages in a safe way (with the tools doing the bookkeeping necessary to avoid dangling edges). Or a tool can provide a higher level model the underlying web structures, thus encouraging authors to give a better structure to their hypertexts, and the readability of the hypertext (for more ideas on hypertext structuring see e.g. [BRS92], see e.g. [Mis95] for information on hypermedia application development). For the development of such tools, a more appropriate data model for the WWW is needed, both for the specification of the internal data structures and data transformations, and for the representation of data that the tool would offer to the user. We believe that graphs and graph grammars are a good candidate for such a data model. In fact: Graphs allow a natural representation of web structures, where the nodes represent units of information (typically pages), and the edges represent hyperlinks. Graph grammars allow to specify complex transformations on graphs in a simple, intuitive way. Recent developments in graph grammars consider higher level structuring of graphs, among these hierarchical structuring (see e.g. [ES95], [Bey93], [BEMW]).